Insert Title Here
Angelos Amyntas1 \(^1\), Emilio Berti \(^1\), Benoit Gauzens \(^1\), Georg Albert \(^1\), […], Nico Eisenhauer \(^2\) & Ulrich Brose \(^1\)
\(^1\)Friederich Schille University of
Jena, \(^2\)University of
Leipzig
\(^1\)Friederich Schille University of Jena, \(^2\)University of Leipzig
08 August, 2022
Abstract
Species-rich communities have shown increased ecosystem functioning such as higher primary productivity or more effective control of herbivores by predator. Additionally, experiments demonstrated that the positive relationship between diversity and ecosystem functioning tends to become steeper over time. One mechanism proposed to explain this change is the reduction of niche overlap within producers and consumers as communities transition to later successional stages. The resulting decrease in competition is hypothesized to be one of the key drivers behind the strengthening of the diversity-functioning relationship. Here we examine how niche differentiation in consumer and producer communities would affect the plant diversity-productivity relationship by combining a food-web assembly model with a bio-energetic model of community dynamics. Starting from a regional pool of species and their potential interactions, we constructed viable local communities that varied in plant richness and niche overlap among species, i.e. with unrestricted or restricted linkage similarity among consumers, crossed with high or low interspecific competition among plants. Then we simulated community dynamics to examine how the different scenarios modify the plant diversity-productivity relationship. Reduced competition had generally minor effects on primary productivity compared to the effect of plant richness. Moreover, we found that a reduction of similarity among consumers did lead to marginally higher productivity in plant-rich and lower in plant-poor communities on average but there was considerable variation in the sign and strength of this relationship and this was more pronounced in communities with fewer plant species. Contrary to our expectations, the reduction of inter-specific competition among plants resulted on average in lower primary productivity across the plant richness gradient. Deciphering the role of competition in ecosystem processes will inform our understanding of how community change over time can modify the functioning of ecosystems which in turn has implications for conservation versus restoration decisions.Introduction
[soon]
Methods
Conseptual figure
Regional species pool
We generated a regional species pool of 1000 species (250 each of plants, herbivores, omnivores, predators). Body-masses range from \(10^{-9}\)g to \(10^3\)g, representing soil organisms from nematodes to small mammals. We arranged herbivorous interactions, including herbivory by omnivores, to have a nested pattern with some plants consumed by most plant consumers, while others by only few and some plant consumers consuming most plant species while others only few of them. Plant consumers were randomly selected across the body size range, then omnivores were sampled from plant consumers with a probability that depends on the fraction of plants that a species consumes (so herbivores are more likely than omnivores to be specialists). Predatory interactions are allometric, following (Schneider et al. 2016) but with an optimal predator prey mass ratio of 0.6 on the log scale, which is typical of terrestrial invertebrates (Brose et al. 2006). The resulting interaction matrix was thinned by randomly removing 30% of the produced interactions. This produced a meta-foodweb where allometry does not absolutely determine consumption.
Local communities
Early succession
Local communities of 2, 4, 8 or 16 plant species and 40 animal species (1000 for each plant diversity level) are produced by sampling the regional species pool while ensuring that all species have at least one resource in the local community (following Bauer et al. (2022)). The produced food-webs represent communities in early successional stages, where species occurence is most contingent on the presence of their resources.
Late succession
For each of the “early succession” communities, a “late succession” community is produced by removing species with a probability that depends on their linkage similarity to other local species (operating as a proxy of niche overlap) and replacing them with species from the regional pool. This proceeds iteratively, with the use of a Metropolis-Hastings optimization algorithm, so that eventually we arrive at a community composition of reduced similarity among the local species (Bauer et al. 2022), representative of communities of late successional stages, where competitive exclusion is at play.
Plant competition
To simulate the effects of niche differentiation on plants, we
directly manipulated the plant-plant interaction matrix to produce
communities where the inter-specific competition among plant
species is strong (early succession) or weaker (late
succession).
For each local community, we first generated a plant competition matrix that corresponds to high niche overlap among species i.e. the competition that each species experiences from all other species can be as high as the competition from con-specifics (fig.). Then we generated a competition matrix that corresponds to reduced niche overlap but without niche expansion i.e inter-specific competition is reduced but intra-specific competition increases such that the overall competition remains the same. Finally, we generated a competition matrix that corresponds to reduced niche overlap combined with niche expansion such that species compete less with other species as well as con-specifics.
16000 food-webs,
Community dynamics
We simulated community dynamics with a bio-energetic model (Delmas et al. 2017; Gauzens and Berti 2022). Changes in the biomass of plants over time are described by the equation
\[\frac{dB_i}{dt} = (r_iG_i-X_i)B_i - \sum_{j}B_jF_{ij}\]
The first term in equation (1) describes biomass gains through growth, where \(B_i\) is the biomass of species \(i\), \(r_i\) is the mass-specific maximum growth rate, \(G_i\) is the net growth rate and \(X_i\) is the mass-specific metabolic loss. The second term describes losses to consumption where \(B_j\) is the biomass of consumer \(j\) and \(F_{ji}\) is the per unit biomass feeding rate of species \(j\) on species \(i\).
The net growth rate of species \(i\) is defined as
\[G_i = 1 - \frac{s_i}{K_i}\]
where \(K_i\) is the carrying capacity of species \(i\). \(s_i\) depends on the ratio of inter- to intra-specific competition for resources \(a_{ij}\).
\[s_i = \sum_{j}a_{ij}B_j \]
The diagonal elements of matrix \(a\) correspond to intraspecific competition while the off-diagonals to interspecific competition.
It is by manipulating the relative strengths of the \(a_{ij}\) elements that we implemented the different plant competition scenarios described above. While restricting row sums to 1, we set \(a_{ij} \leq a_{ii}\) for the baseline scenario, \(a_{ij} << a_{ii}\) for the shift to higher intra-specific competition. Finally, combining the \(a_{ii}\) values of the former scenario with the \(a_{ij}\) values of the later scenario, we produced communities with an overall decrease in plant competition. In all cases the \(a\) matrix was multiplied by \(N\) (number of plant species in the community) such that the overall competition experienced by one species would sum to \(N\) for the first two competition scenarios, or to a fraction of \(N\) for the reduced competition scenario.
Changes in the biomass of animals over time are described by the equation
\[\frac{dB_i}{dt} = B_i(\sum_{j}F_{ji}e_{j} - X_i) - \sum_{j}B_jF_{ij}\]
The first term in equation () describes biomass gains through consumption (after metabolic loss?) where \(B_i\) is the biomass of species \(i\), \(F_{ji}\) is the per unit biomass feeding rate of species \(i\) on species \(j\), \(e_j\) is the assimilation efficiency of prey \(j\). The second term describes losses to consumption as in eq. 1.
Results
Out of 24000 simulated communities, 516 (2.15%) had no plant species surviving at the end of simulated dynamics and were discarded. Communities with a high initial plant richness had on average a larger proportion of plant and animal species surviving at the end of simulations (supplement?). This resulted in a positive plant richness - animal richness relationship.
The simulated communities reproduced the expected positive and
saturating plant richness-productivity relationship (fig. 1), as well as
a positive relationship between richness and standing plant biomass
(figSA). When comparing our scenarios of competition change to a
baseline of unrestricted similarity among animals and high
inter-specific as well as intra-specific competition among plants
(fig. 1a), we found that a reduction of similarity (and therefore
competition) among animals had a marginally positive
practically no effect in the richness-productivity relationship (fig1b)
(but had a more pronounced positive effect on the richness-biomass
relationship (fig.SAb)).
Changes in plant competition had generally stronger effects. When the change in competition resulted in plant species competing less with other species but stronger with con-specifics, there was on average a negative effect on the richness-productivity relationship across all richness levels (although the effect was variable at low plant richness and very consistent at high plant richness) (fig. 1c). By contrast, lowering competition overall in the plant community led to a consistent relative increase in productivity at higher richness levels (fig. 1d), corresponding to a steepening of the richness-diversity slope.
Finally, the combination of reduced animal and plant competition had on average a positive difference in productivity that was larger at higher plant richness (fig. 1e). Additionally, the difference in productivity of communities with low versus high animal competition when plant competition is low (fig. SB), shows that the reduction in animal competition does have a small contribution to the relationship shown in fig.1e.
Figure 1. Primary productivity at equilibrium. (a) Primary productivity at different levels of initial plant richness for communities with unrestricted animal similarity and high interspecific plant competition (a). The difference in productivity relative to (a) when animals are less similar to each other (b), when plants compete less with other species but stronger with conspecifics (c), when plants compete less overall (d), b & d combined (e). Black points are mean values, thick black lines represent the 75% density interval, thin black lines the 95% density interval.
\(~\)
The pressure of herbivory on plants (quantified as total out-fluxes from plants to their consumers, including omnivores, divided by plant biomass) was lower at higher richness levels (fig2). The difference in herbivory pressure as a result of changes in animal or plant competition was highly variable at low plant richness and diminished at high richness but, overall, competition changes did not have an effect on the slope or intercept of the richness-herbivory-pressure relationship (figs 2b-d).
Figure 2. Herbivory pressure at equilibrium (outfluxes from plants to plant consumers incl. omnivores divided by plant biomass). (a) Herbivory pressure at different levels of initial plant richness for communities with unrestricted animal similarity and high interspecific plant competition (a). The difference in productivity relative to (a) when animals are less similar to each other (b), plants compete less with other species but stronger with conspecifics (c), plants compete less overall (d), b & d combined (e). Black points are mean values, thick black lines represent the 75% density interval, thin black lines the 95% density interval.
The control of herbivores by predators (quantified as the ratio of out-fluxes to in-fluxes of herbivores) had a hump-shaped relationship with plant richness (fig.3). This was largely the result of no control at the lowest richness level as a majority of the 2 plant species communities consisted of plants and herbivores alone with no higher levels surviving. A reduction in animal competition resulted in a relative increase in herbivory-control on average, that was more pronounced at intermediate richness levels (fig. 3b). A shift of plant competition towards stronger intra-specific competition had the opposite effect (fig. 3c), corresponding to a dampening of the hump-shaped relationship. Conversely, lower plant competition resulted in a positive difference in herbivore-control at higher plant richness (fig. 3d). Finally the combined change of lower animal and plant competition, also had a positive difference in herbivore control at higher plant richness. Again the contribution of a reduction of animal competition is clear when comparing the differene in herbivore control among communities with low versus high animal competition when plant competition is low (fig. SC).
